Table 2 Overall and subgroup analyses for the association between selenium levels and the odds of metabolic syndrome
Hormones | Subgroups | Studies (effect sizes) | Test of effect | Test of heterogeneity | |
|---|---|---|---|---|---|
OR (95%CI) | I2 (%) | P | |||
Overall |  | 18 (26) | 1.10 (1.03–1.17) | 87.3 | 0.001 |
Population | Community-based | 14 (20) | 1.09 (0.95–1.25) | 86.3 | 0.001 |
|  | Hospital-based | 4 (6) | 2.49 (1.57–3.96) | 73.7 | 0.002 |
Quality of studies | Low risk of bias | 13 (21) | 1.36 (1.12–1.66) | 88.2 | 0.001 |
|  | High risk of bias | 5 (5) | 1.29 (0.86–1.99) | 92.8 | 0.001 |
Gender of participants | Both | 12 (12) | 1.02 (0.96–1.08) | 85.6 | 0.001 |
|  | Male | 8 (8) | 1.32 (0.96–1.81) | 76.7 | 0.001 |
|  | Female | 6 (6) | 2.00 (1.17–3.43) | 94.2 | 0.001 |
MetS definition | NCEP-ATP III | 5 (7) | 1.75 (1.10–2.80) | 86.9 | 0.001 |
|  | IDF | 5 (6) | 1.25 (0.91 to 1.72) | 91.8 | < 0.000 |
|  | Chinese criteria | 5 (10) | 1.69 (1.14 to 2.50) | 82.4 | < 0.000 |
|  | JIS | 1 (1) | 0.61 (0.45 to 0.83) | 0.0 | < 0.000 |
|  | NR | 2 (2) | 1.00 (0.37 to 2.71) | 92.1 | < 0.000 |
Sample size | ≥ 1000 participants | 9 (12) | 1.16 (0.90 to 1.50) | 84.6 | < 0.000 |
|  | < 1000 participants | 9 (14) | 1.65 (1.34 to 2.05) | 89.8 | < 0.000 |
Age of participant | ≥ 50 years | 12 (19) | 1.64 (1.23 to 2.20) | 89.5 | < 0.000 |
|  | < 50 years | 4 (5) | 1.10 (0.94 to 1.28) | 70.3 | 0.009 |
| Â | NR | 2 (2) | 0.82 (0.39 to 1.72) | 65.1 | 0.090 |
Study type | Case-control | 9 (16) | 1.66 (1.33 to 2.06) | 91.5 | < 0.000 |
|  | Cohort | 1 (1) | 1.12 (0.88 to 1.42) | 0.0 | < 0.000 |
| Â | Cross- sectional | 8 (9) | 1.09 (0.88 to 1.36) | 69.4 | 0.001 |
Geographical region | Asian | 16 (23) | 1.52 (1.21 to 1.90) | 88.7 | < 0.000 |
| Â | Non- Asian | 2 (3) | 1.05 (0.88 to 1.25) | 72.7 | 0.026 |
Selenium assessment | ICP-MS | 11 (15) | 1.41 (1.03 to 1.93) | 86.0 | < 0.000 |
|  | Atomic absorption spectroscopy | 4 (8) | 1.80 (1.30 to 2.49) | 91.2 | < 0.000 |
|  | Spectrofluorimetry | 1 (1) | 0.95 (0.87 to 1.04) | 0.0 | < 0.000 |
|  | Neutron activation analysis | 1 (1) | 1.33 (0.58 to 3.05) | 0.0 | < 0.000 |
|  | NR | 1 (1) | 0.61 (0.43 to 0.87) | 0.0 | < 0.000 |
Sample type | Serum | 15 (23) | 1.43 (1.22 to 1.68) | 89.0 | < 0.000 |
| Â | Urine | 2 (2) | 0.89 (0.70 to 1.14) | 0.0 | 0.469 |
|  | Toenail | 1 (1) | 1.33 (0.58 to 3.05) | 0.0 | < 0.000 |
Adjustment for any covariates | Yes | 15 (23) | 1.42 (1.19 to 1.70) | 85.4 | < 0.000 |
|  | No | 3 (3) | 1.27 (0.58 to 2.82) | 95.9 | < 0.000 |
Adjustment for age | Yes | 13 (20) | 1.26 (1.08 to 1.47) | 78.5 | < 0.000 |
|  | No | 5 (6) | 1.94 (0.92 to 4.10) | 96.1 | < 0.000 |
Adjustment for sex | Yes | 10 (16) | 1.39 (1.09 to 1.77) | 81.0 | < 0.000 |
|  | No | 8 (10) | 1.46 (1.09 to 1.96) | 92.1 | < 0.000 |
Adjustment for smoking | Yes | 11 (17) | 1.44 (1.11 to 1.87) | 85.4 | < 0.000 |
|  | No | 7 (9) | 1.27 (1.03 to 1.56) | 89.0 | < 0.000 |
Adjustment for alcohol drinking | Yes | 11 (17) | 1.44 (1.11 to 1.87) | 85.4 | < 0.000 |
|  | No | 7 (9) | 1.27 (1.03 to 1.56) | 89.0 | < 0.000 |
Adjustment for total energy intake | Yes | 2 (3) | 1.14 (0.88 to 1.48) | 58.1 | 0.11 |
|  | No | 16 (23) | 1.42 (1.20 to 1.67) | 88.7 | < 0.000 |
Adjustment for BMI | Yes | 5 (7) | 1.42 (0.80 to 2.50) | 90.0 | < 0.000 |
|  | No | 13 (19) | 1.38 (1.18 to 1.61) | 87.4 | < 0.000 |
Adjustment for education | Yes | 5 (5) | 0.93 (0.65 to 1.32) | 71.3 | 0.10 |
|  | No | 13 (21) | 1.49 (1.26 to 1.75) | 89.0 | < 0.000 |
Adjustment for physical activity | Yes | 8 (12) | 1.63 (1.15 to 2.30) | 89.7 | 0.001 |
| Â | No | 10 (14) | 1.22 (1.03 to 1.43) | 83.7 | 0.001 |